This section describes background subject matter related to the disclosed embodiments of the present invention. There is no intention, either express or implied, that the background art discussed in this section legally constitutes prior art.
Nanostructuring is necessary for many present applications and industries and for new technologies which are under development. Improvements in efficiency can be achieved for current applications in areas such as solar cells and LEDs, and in next generation data storage devices, for example and not by way of limitation.
Nanostructured substrates may be fabricated using techniques such as e-beam direct writing, Deep UV lithography, nanosphere lithography, nanoimprint lithography, near-filed phase shift lithography, and plasmonic lithography, for example.
NanoImprint Lithography (NIL) creates patterns by mechanical deformation of an imprint resist, followed by subsequent processing. The imprint resist is typically a monomeric or polymeric formulation that is cured by heat or by UV light during the imprinting. There are a number of variations of NIL. However, two of the processes appear to be the most important. These are Thermoplastic NanoImprint Lithography (TNIL) and Step and Flash NanoImprint Lithography (SFIL).
TNIL is the earliest and most mature nanoimprint lithography. In a standard TNIL process, a thin layer of imprint resist (a thermoplastic polymer) is spin coated onto a sample substrate. Then a mold, which has predefined topological patterns, is brought into contact with the sample, and pressed against the sample under a given pressure. When heated above the glass transition temperature of the thermoplastic polymer, the pattern on the mold is pressed into a thermoplastic polymer film melt. After the sample, with impressed mold is cooled down, the mold is separated from the sample and the imprint resist is left on the sample substrate surface. The pattern does not pass through the imprint resist; there is a residual thickness of unchanged thermoplastic polymer film remaining on the sample substrate surface. A pattern transfer process, such as reactive ion etching, can be used to transfer the pattern in the resist to the underlying substrate. The variation in the residual thickness of unaltered thermoplastic polymer film presents a problem with respect to uniformity and optimization of the etch process used to transfer the pattern to the substrate.
Tapio Makela et al. of VTT, a technical research center in Finland, have published information about a custom built laboratory scale roll-to-roll imprinting tool dedicated to manufacturing of submicron structures with high throughput. Hitachi and others have developed a sheet or roll-to-roll prototype NIL machine, and have demonstrated capability to process 15 meter long sheets. The goal has been to create a continuous imprint process using belt molding (nickel plated molds) to imprint polystyrene sheets for large geometry applications such as membranes for fuel cells, batteries and possibly displays.
Hua Tan et al of Princeton University have published 2 implementations of roller Nanoimprint lithography: rolling cylinder mold on flat, solid substrate, and putting a flat mold directly on a substrate and rolling a smooth roller on top of the mold. Both methods are based on TNIL approach, where roller temperature is set above the glass transition temperature, Tg, of the resist (PMMA), while the platform is set to temperature below Tg. Currently the prototype tools do not offer a desirable throughput. In addition, there is a need to improve reliability and repeatability with respect to the imprinted surface.
In the SFIL process, a UV curable liquid resist is applied to the sample substrate and the mold is made of a transparent substrate, such as fused silica After the mold and the sample substrate are pressed together, the resist is cured using UV light, and becomes solid. After separation of the mold from the cured resist material, a similar pattern to that used in TNIL may be used to transfer the pattern to the underlying sample substrate. Dae-Geun Choi from Korea Institute of Machinery suggested using fluorinated organic-inorganic hybrid mold as a stamp for Nanoimprint lithography, which does not require anti-stiction layer for demolding it from the substrate materials.
Since Nanoimprint lithography is based on mechanical deformation of resist, there are a number of challenges with both the SFIL and TNIL processes, in static, step-and-repeat, or roll-to-roll implementations. Those challenges include template lifetime, throughput rate, imprint layer tolerances, and critical dimension control during transfer of the pattern to the underlying substrate. The residual, non-imprinted layer which remains after the imprinting process requires an additional etch step prior to the main pattern transfer etch. Defects can be produced by incomplete filling of negative patterns and the shrinkage phenomenon which often occurs with respect to polymeric materials. Difference in thermal expansion coefficients between the mold and the substrate cause lateral strain, and the strain is concentrated at the corner of the pattern. The strain induces defects and causes fracture defects at the base part of the pattern mold releasing step.
Soft lithography is an alternative to Nanoimprint lithography method of micro and nano fabrication. This technology relates to replica molding of self assembling monolayers. In soft lithography, an elastomeric stamp with patterned relief structures on its surface is used to generate patterns and structures with feature sizes ranging from 30 nm to 100 nm. The most promising soft lithography technique is microcontact printing (μCP) with self-assembled monolayers (SAMS). The basic process of μCP includes: 1. A polydimethylsiloxane (PDMS) mold is dipped into a solution of a specific material, where the specific material is capable of forming a self-assembled monolayer (SAM). Such specific materials may be referred to as an ink. The specific material sticks to a protruding pattern on the PDMS master surface. 2. The PDMS mold, with the material-coated surface facing downward, is contacted with a surface of a metal-coated substrate such as gold or silver, so that only the pattern on the PDMS mold surface contacts the metal-coated substrate. 3. The specific material forms a chemical bond with the metal, so that only the specific material which is on the protruding pattern surface sill remain on the metal-coated surface after removal of the PDMS mold. The specific material forms a SAM on the metal-coated substrate which extends above the metal-coated surface approximately one to two nanometers (just like ink on a piece of paper). 4. The PDMS mold is removed from the metal-coated surface of the substrate, leaving the patterned SAM on the metal-coated surface.
The best-established specific materials for forming SAMs on gold or silver-coated surfaces are alkanethiolates. When the substrate surface contains hydroxyl-terminated moieties such as Si/SiO2, Al/Al2O3, glass, mica, and plasma-treated polymers, alkylsiloxanes work well as the specific materials. With respect to the alkanethiolates, μCP of hexadecanethiol on evaporated thin (10-200 nm thick) films of gold or silver appears to be the most reproducible process. While these are the best-known materials for carrying out the pattern formation, gold and silver are not compatible with microelectronic devices based on silicon technology, although gold or silver-containing electrodes or conductive wires may used. Currently, μCP for SAMS of siloxanes on Si/SiO2 surfaces are not as tractable as the SAMS of alkanethiolates on gold or silver. The SAMS of siloxanes on Si/SiO2 often provide disordered SAMs, and in some cases generate submonolayers or multilayers. Finally, the patterned molds available for μCP are flat “stamp” surfaces, and reproducible and reliable printing on large areas not only requires very accurate stitching of the printed pattern from the mold, but also requires constant wetting of the stamp with the SAM-forming specific material, which is quite problematic.
Optical Lithography does not use mechanical deformation or phase change of resist materials, like Nanoimprint lithography, and does not have materials management problems like Soft Lithography, thus providing better feature replication accuracy and more Manufacturable processing. Though regular optical lithography is limited in resolution by diffraction effects some new optical lithography techniques based on near field evanescent effects have already demonstrated advantages in printing sub-100 nm structures, though on small areas only. Near-field phase shift lithography NFPSL involves exposure of a photoresist layer to ultraviolet (UV) light that passes through an elastomeric phase mask while the mask is in conformal contact with a photoresist. Bringing an elastomeric phase mask into contact with a thin layer of photoresist causes the photoresist to “wet” the surface of the contact surface of the mask. Passing UV light through the mask while it is in contact with the photoresist exposes the photoresist to the distribution of light intensity that develops at the surface of the mask. In the case of a mask with a depth of relief that is designed to modulate the phase of the transmitted light by π, a local null in the intensity appears at the step edge of relief. When a positive photoresist is used, exposure through such a mask, followed by development, yields a line of photoresist with a width equal to the characteristic width of the null in intensity. For 365 nm (Near UV) light in combination with a conventional photoresist, the width o the null in intensity is approximately 100 nm. A PDMS mask can be used to form a conformal, atomic scale contact with a flat, solid layer of photoresist. This contact is established spontaneously upon contact, without applied pressure. Generalized adhesion forces guide this process and provide a simple and convenient method of aligning the mask in angle and position in the direction normal to the photoresist surface, to establish perfect contact. There is no physical gap with respect to the photoresist. PDMS is transparent to UV light with wavelengths greater than 300 nm. Passing light from a mercury lamp (where the main spectral lines are at 355-365 nm) through the PDMS while it is in conformal contact with a layer of photoresist exposes the photoresist to the intensity distribution that forms at the mask.
Yasuhisa Inao, in a presentation entitled “Near-Field Lithography as a prototype nano-fabrication tool”, at the 32nd International Conference on Micro and Nano Engineering in 2006, described a step-and-repeat near-field nanolithography developed by Canon, Inc. Near-field lithography (NFL) is used, where the distance between a mask and the photoresist to which a pattern is to be transferred are as close as possible. The initial distance between the mask and a wafer substrate was set at about 50 μm. The patterning technique was described as a “tri-layer resist process”, using a very thin photoresist. A pattern transfer mask was attached to the bottom of a pressure vessel and pressurized to accomplish a “perfect physical contact” between the mask and a wafer surface. The mask was “deformed to fit to the wafer”. The initial 50 μm distance between the mask and the wafer is said to allows movement of the mask to another position for exposure and patterning of areas more than 5 mm×5 mm. The patterning system made use of i-line (365 nm) radiation from a mercury lamp as a light source. A successful patterning of a 4 inch silicon wafer with structures smaller than 50 nm was accomplished by such a step-and-repeat method.
In an article entitled “Large-area patterning of 50 nm structures on flexible substrates using near-field 193 nm radiation”, JVST B 21 (2002), at pages 78-81, Kunz et al. applied near-field phase shift mask lithography to the nanopatterning of flexible sheets (Polyimide films) using rigid fused silica masks and deep UV wavelength exposure. In a subsequent article entitled “Experimental and computational studies of phase shift lithography with binary elastomeric masks”, JVST B 24(2) (2006) at pages 828-835, Maria et al. present experimental and computational studies of a phase shifting photolithographic technique that uses binary elastomeric phase masks in conformal contact with layers of photoresist. The work incorporates optimized masks formed by casting and curing prepolymers to the elastomer poly(dimethylsiloxane) against anisotropically etched structures of single crystal silicon on SiO2/Si. The authors report on the capability of using the PDMS phase mask to form resist features in the overall geometry of the relief on the mask.
U.S. Pat. No. 6,753,131 to Rogers et al, issued Jun. 22, 2004, titled “Transparent Elastomeric, Contact-Mode Photolithography Mask, Sensor, and Wavefront Engineering Element”, describes a contact-mode photolithography phase mask which includes a diffracting surface having a plurality of indentations and protrusions. The protrusions are brought into contact with a surface of a positive photoresist, and the surface is exposed o electromagnetic radiation through the phase mask. The phase shift due to radiation passing through indentations as opposed to the protrusions is essentially complete. Minima in intensity of electromagnetic radiation are thereby produced at boundaries between the indentations and protrusions. The elastomeric mask conforms well to the surface of the photoresist, and following development of the photoresist, features smaller than 100 nm can be obtained. (Abstract) In one embodiment, reflective plates are used exterior to the substrate and the contact mask, so radiation will be bounced to a desired location at a shifted phase. In another embodiment, the substrate may be shaped in a manner which causes a deformation of the phase shifting mask, affecting the behavior of the phase shifting mask during exposure.
Near Field Surface Plasmon Lithography (NFSPL) makes use of near-field excitation to induce photochemical or photophysical changes to produce nanostructures. The main near-field technique is based on the local field enhancement around metal nanostructures when illuminated at the surface plasmon resonance frequency. Plasmon printing consists of the use of plasmon guided evanescent waves through metallic nanostructures to produce photochemical and photophysical changes in a layer below the metallic structure. In particular, visible exposure (λ=410 nm) of silver nanoparticles in close proximity to a thin film of a g-line photoresist (AZ-1813 available from AZ-Electronic Materials, MicroChemicals GmbH, Ulm, Germany) can produce selectively exposed areas with a diameter smaller than λ/20. W. Srituravanich et al. in an article entitled “Plasmonic Nanolithography”, Nanoletters V4, N6 (2004), pp. 1085-1088, describes the use of near UV light (λ=230 nm-350 nm) to excite SPs on a metal substrate, to enhance the transmission through subwavelength periodic apertures with effectively shorter wavelengths compared to the excitation light wavelength. A plasmonic mask designed for lithography in the UV range is composed of an aluminum layer perforated with 2 dimensional periodic hole arrays and two surrounding dielectric layers, one on each side. Aluminum is chosen since it can excite the SPs in the UV range. Quartz is employed as the mask support substrate, with a poly(methyl methacrylate) spacer layer which acts as adhesive for the aluminum foil and as a dielectric between the aluminum and the quartz. Poly(methyl methacrylate is used in combination with quartz, because their transparency to UV light at the exposure wavelength (i-line at 365 nm) and comparable dielectric constants (2.18 and 2.30, quartz and PMMA, respectively). A sub-100 nm dot array pattern on a 170 nm period has been successfully generated using an exposure radiation of 365 nm wavelength. Apparently the total area of patterning was about 5 μm×5 μm, with no scalability issues discussed in the paper.
Joseph Martin has suggested a proximity masking device for Near-filed lithography in U.S. Pat. No. 5,928,815, where cylindrical block covered with metal film for light internal reflection is used for directing light to the one end of the cylinder (base of the cylinder), which contains a surface relief pattern used for Near-field exposure. This block is kept in some proximity distance (“very small, but not zero”) from the photoresist on the sample. Cylinder is translated in horizontal direction using some precise mechanism, which is used to pattern photoresist area.
The only published idea about using rollers for optical lithography can be found in the Japanese Unexamined Patent Publication, No. 59200419A, published Nov. 13, 1984, titled “Large Area Exposure Apparatus”. Toshio Aoki et al. described the use of a transparent cylindrical drum which can rotate and translate with an internal light source and a film of patterned photomask material attached on the outside of the cylindrical drum. A film of a transparent heat reflective material is present on the inside of the drum. A substrate with an aluminum film on its surface and a photoresist overlying the aluminum film is contacted with the patterned photomask on the drum surface and imaging light is passed through the photomask to image the photoresist on the surface of the aluminum film. The photoresist is subsequently developed, to provide a patterned photoresist. The patterned photoresist is then used as an etch mask for an aluminum film present on the substrate.
There is no description regarding the kinds of materials which were used as a photomask film or as a photoresist on the surface of the aluminum film. A high pressure mercury lamp light source (500 W) was used to image the photoresist overlying the aluminum film. Glass substrates about 210 mm (8.3 in.)×150 mm (5.9 in.) and about 0.2 mm (0.008 in.) thick were produced using the cylindrical drum pattern transfer apparatus. The feature size of the pattern transferred using the technique was about 500 μm2, which was apparently a square having a dimension of about 22.2 μm×22.2 μm. This feature size was based on the approximate pixel size of an LCD display at the time the patent application was filed in 1984. The photomask film on the outside of the cylindrical drum was said to last for approximately 140,000 pattern transfers. The contact lithography scheme used by Toshio Aoki et al. is not capable of producing sub-micron features.
It does not appear that an nanoimprinting methods (thermal or UV-cured) or soft lithography using printing with SAM materials are highly manufacturable processes. In general, the imprinting method creates deformation of the substrate material due to the thermal treatment (thermal NIL, for example) or shrinkage of pattern features upon polymer curing (UV-cured polymeric features). Moreover, due to the application of pressure (hard contact) between a stamp and a substrate, defects are essentially unavoidable, and a stamp has a very limited lifetime. Soft lithography does have an advantage in that it is thermal and stress-free printing technology. However, the use of a SAM as an “ink” for a sub-100 nm pattern is very problematic due to the drifting of molecules over the surface, and application over large areas has not been proved experimentally.